rust/tantivy
1
0
Fork 0
mirror of https://github.com/quickwit-oss/tantivy.git synced 2025-12-23 02:29:57 +00:00
Code Issues Packages Projects Releases Wiki Activity

support ff range queries on json fields (#2456)

Browse Source 
* support ff range queries on json fields

* fix term date truncation

* use inverted index range query for phrase prefix queries

* rename to InvertedIndexRangeQuery

* fix column filter, add mixed column test
This commit is contained in:
PSeitz
2024-08-02 01:06:50 +09:00
committed by GitHub
parent 0d4e319965
commit 3d1c4b313a
15 changed files with 1322 additions and 289 deletions
Download Patch File Download Diff File Expand all files Collapse all files

50
src/query/query_parser/query_parser.rs
View File

@@ -137,7 +137,7 @@ fn trim_ast(logical_ast: LogicalAst) -> Option<LogicalAst> {
/// so-called default fields (as set up in the constructor).
///
/// Assuming that the default fields are `body` and `title`, and the query parser is set with
/// conjunction as a default, our query will be interpreted as.
/// conjunction as a default, our query will be interpreted as.
/// `(body:Barack OR title:Barack) AND (title:Obama OR body:Obama)`.
/// By default, all tokenized and indexed fields are default fields.
///
@@ -148,8 +148,7 @@ fn trim_ast(logical_ast: LogicalAst) -> Option<LogicalAst> {
/// `body:Barack OR (body:Barack OR text:Obama)` .
///
/// * boolean operators `AND`, `OR`. `AND` takes precedence over `OR`, so that `a AND b OR c` is
/// interpreted
/// as `(a AND b) OR c`.
/// interpreted as `(a AND b) OR c`.
///
/// * In addition to the boolean operators, the `-`, `+` can help define. These operators are
/// sufficient to express all queries using boolean operators. For instance `x AND y OR z` can be
@@ -272,8 +271,7 @@ impl QueryParser {
/// Creates a `QueryParser`, given
/// * an index
/// * a set of default fields used to search if no field is specifically defined
/// in the query.
/// * a set of default fields used to search if no field is specifically defined in the query.
pub fn for_index(index: &Index, default_fields: Vec<Field>) -> QueryParser {
QueryParser::new(index.schema(), default_fields, index.tokenizers().clone())
}
@@ -482,16 +480,33 @@ impl QueryParser {
});
if terms.len() != 1 {
return Err(QueryParserError::UnsupportedQuery(format!(
"Range query boundary cannot have multiple tokens: {phrase:?}."
"Range query boundary cannot have multiple tokens: {phrase:?} [{terms:?}]."
)));
}
Ok(terms.into_iter().next().unwrap())
}
FieldType::JsonObject(_) => {
// Json range are not supported.
Err(QueryParserError::UnsupportedQuery(
"Range query are not supported on json field.".to_string(),
))
FieldType::JsonObject(ref json_options) => {
let get_term_with_path = || {
Term::from_field_json_path(
field,
json_path,
json_options.is_expand_dots_enabled(),
)
};
if let Some(term) =
// Try to convert the phrase to a fast value
convert_to_fast_value_and_append_to_json_term(
get_term_with_path(),
phrase,
false,
)
{
Ok(term)
} else {
let mut term = get_term_with_path();
term.append_type_and_str(phrase);
Ok(term)
}
}
FieldType::Facet(_) => match Facet::from_text(phrase) {
Ok(facet) => Ok(Term::from_facet(field, &facet)),
@@ -553,7 +568,7 @@ impl QueryParser {
}
FieldType::Date(_) => {
let dt = OffsetDateTime::parse(phrase, &Rfc3339)?;
let dt_term = Term::from_field_date(field, DateTime::from_utc(dt));
let dt_term = Term::from_field_date_for_search(field, DateTime::from_utc(dt));
Ok(vec![LogicalLiteral::Term(dt_term)])
}
FieldType::Str(ref str_options) => {
@@ -685,8 +700,8 @@ impl QueryParser {
///
/// The terms are identified by a triplet:
/// - tantivy field
/// - field_path: tantivy has JSON fields. It is possible to target a member of a JSON
/// object by naturally extending the json field name with a "." separated field_path
/// - field_path: tantivy has JSON fields. It is possible to target a member of a JSON object by
/// naturally extending the json field name with a "." separated field_path
/// - field_phrase: the phrase that is being searched.
///
/// The literal identifies the targeted field by a so-called *full field path*,
@@ -949,7 +964,8 @@ fn generate_literals_for_json_object(
|| Term::from_field_json_path(field, json_path, json_options.is_expand_dots_enabled());
// Try to convert the phrase to a fast value
if let Some(term) = convert_to_fast_value_and_append_to_json_term(get_term_with_path(), phrase)
if let Some(term) =
convert_to_fast_value_and_append_to_json_term(get_term_with_path(), phrase, true)
{
logical_literals.push(LogicalLiteral::Term(term));
}
@@ -1123,8 +1139,8 @@ mod test {
let query = make_query_parser().parse_query("title:[A TO B]").unwrap();
assert_eq!(
format!("{query:?}"),
"RangeQuery { lower_bound: Included(Term(field=0, type=Str, \"a\")), upper_bound: \
Included(Term(field=0, type=Str, \"b\")), limit: None }"
"RangeQuery { bounds: BoundsRange { lower_bound: Included(Term(field=0, type=Str, \
\"a\")), upper_bound: Included(Term(field=0, type=Str, \"b\")) } }"
);
}